USGS visual identity mark and link to main Web site

USGS Open-File Report 94-023

GCM Simulations Of The Pliocene Climate: Feedbacks, Ocean Transports, And CO2

Mark Chandler
NASA/GISS, New York, New York 10025
Estimates of sea surface temperatures (SSTs), based on marine microfauna, reveal the existence of a middle Pliocene warm period between about 3.15 and 2.85 million years ago (Dowsett et al., 1992). Terrestrial pollen records, although not as well dated, also show evidence for a warmer climate at about this same time in the Pliocene and, further, indicate that continental moisture levels varied significantly from the present day. The cause of the altered climate is not known with certainty, but sensitivity experiments, conducted using the Goddard Institute for Space Studies General Circulation Model (GISS GCM), have indicated that warmer climates, such as those of the Pliocene, can be simulated by using increased ocean heat transports (Rind and Chandler, 1991) [figure 1]. Dowsett et al. (1992) suggested that this might be the case for the Pliocene, based on the distribution of North Atlantic SSTs.

Figure 1. Pliocene and modern annual ocean heat transports for the Atlantic Ocean
This figure is available as a GIF, PICT, or TIFF (line-art) image.
One test of this hypothesis is to supply Pliocene SSTs, together with an estimate of the terrestrial vegetation coverage, as boundary conditions in a GCM simulation, and to examine the temperature feedbacks and moisture changes that result. Consistency between the palynological estimates of climate and the simulation results provide one level of validation for the GCM; the GCM then provides a method for investigating the atmospheric processes involved in maintaining the warmer climate.

Using the GISS GCM together with PRISM Northern Hemisphere ocean surface and vegetation boundary conditions (Chandler et al., submitted; Dowsett et al., submitted) we found both consistencies and inconsistencies between model and data-generated paleoclimate estimates. Generally, temperature estimates show the greatest consistency, with both model and data indicating significantly warmer temperatures at high latitudes and diminished warming nearer to the equator. The GCM yields temperature increases up to 10C along the Arctic coasts and shows greatest warming in the winter. Although the original temperature increase is driven by warmer SSTs, much of the continental interior warming is generated by an ice-albedo feedback, as reduced snow cover in the warmer climate increases the absorption of solar radiation at the surface during winter months [figure 2]. Further warming at high latitudes comes from the increased levels of atmospheric water vapor (a greenhouse gas) that are a result of the warm, ice-free ocean conditions. Despite the generally warmer climatic conditions, some areas show overall cooling. Notably, East Africa cools by 2 to 3 C due to increased low-level cloud cover which reflects large amounts of incoming solar radiation back to space. This result is consistent with the single palynological record that exists for that region.

Figure 2. Water vapor, cloud coverage, and ground albedo in the northern hemisphere as a function of latitude in the GISS Pliocene GCM simulation
This figure is available as a GIF, PICT, or TIFF (line-art) image.
Model-data moisture estimates show far less consistency than do the temperature estimates, not a surprising result given the complexities involved in modeling hydrologic processes using coarse-grid numerical models like the GISS GCM. The most common discordance seems to be an underestimation of the increased wetness suggested by pollen records at several localities throughout the Northern Hemisphere. The model's simple ground hydrology responds to the warmer summer ground temperatures by drying out while the diminished intensity of the atmospheric circulation (a result of reduced latitudinal temperature gradients) decreases the amount of moisture advected from over the oceans to the continents. In the Arctic, where modern tundra environments were replaced by Pliocene boreal forests, the altered boundary conditions required that wetter soil moisture conditions be specified. The Pliocene Arctic soils remained wetter than the present day, indicating that the specified wet conditions were in equilibrium with the simulated climate.

In addition to the above experiment, several simulations were conducted using increased levels of atmospheric carbon dioxide; higher CO2 amounts have also been proposed as a potential cause of the warmer Pliocene climates (Crowley, 1991). Rind and Chandler, (1991) pointed out that SST patterns such as the one seen in the Pliocene are inconsistent with CO2 generated warming, however, it is possible that some combination of CO2 increase and ocean heat transport change could have resulted in the warmer Pliocene surface temperatures. Figure 3 shows the various levels of ocean heat transport required to generate the PRISM SSTs given various atmospheric CO2 increases. The graph indicates that with modern ocean heat transports (0% increase) CO2 levels must have been at least 1400 ppm (4.5 times the modern value) in order to generate the global warming of the Pliocene. So far, estimates based on carbon isotope measurements by Raymo and Rau (1992) suggest that Pliocene CO2 levels were, at most, 100 ppm greater than today.

Figure 3. Levels of ocean heat transport required to generate PRISM sea-surface temperatures at varying level of atmospheric CO2 concentrations
This figure is available as a GIF PICT or TIFF (line-art) image.


This page is <>
Maintained by Eastern Publications Group Web Team
Last modified 28-Feb-2001